How to evaluate access to finance – Mexico microcredit programme (randomised control trial)

What was the programme and what did it aim to do?

This study examines the impact of a microcredit programme implemented in a region of Mexico in 2009. The programme was implemented by Compartamos Banco, Mexico’s largest micro-lender, which targets its group lending operations at working-age women who own, or would like to own, a business. The loan amounts ranged from £70-£1,280 and were to be repaid over 16 weeks with an interest rate of about 110% APR. Groups of 10-50 members were jointly liable for repayment of the loan. Since Compartamos is a business, the programme was primarily profit motivated. However, microcredit programmes such as this are often thought to be an effective development tool.

What’s the evaluation challenge?

Evaluating the effect of microcredit programmes is difficult because such programmes are used by specific types of individual. The lender has a target market (in this case working age women) and only a selection of the sample market will be interested in taking up a loan (e.g. those having trouble getting credit elsewhere). The groups also screen new members and their business plans (due to joint liability), meaning only potentially profitable business ventures will get loans. As a result of this selection, if we compare differences in outcomes for individuals who receive loans to those who do not, these differences may not reflect the impact of the programme. Instead, they may simply reflect differences in the types of individuals who receive loans.

What did the evaluation do?

The authors of the study worked with Compartamos to implement a Randomised Control Trial (RCT). The study design was based on randomisation at the neighbourhood rather than individual level (a so called cluster-randomised design). There were 250 neighbourhoods, half of which were randomly assigned to the treatment group. In these neighbourhoods Compartamos sent loan officers to promote their microcredit programme. The remaining control neighbourhoods received no loan promotion. Loan promotion increases the likelihood that individuals in a neighbourhood used the programme (18.9% take-up in treatment areas vs. 5.8% take-up in controls). The study then examined outcomes such as credit access, self-employment, and self-reported welfare for individuals in treatment and control areas. Given the neighbourhoods were randomly assigned, the difference in outcomes may be interpreted as the causal effect of the policy.

How good was the evaluation?

According to our scoring guide, an RCT receives a maximum of 5 (out of 5) on the Maryland Scientific Methods Scale (Maryland SMS). This is because randomisation ensures there are neither observable (e.g. age) nor unobservable differences (e.g. motivation) between treatment and control groups so comparisons likely reflect the impact of the programme. The study is well implemented in a number of ways. For example, it demonstrates randomisation was successful by presenting balancing tests which show the treatment and control groups are statistically similar on a range of characteristics. Therefore, we score the study a maximum 5 on the SMS.

What did the evaluation find?

The study finds that promotion of the microcredit programme leads to increases in overall household borrowing (by £60 from a base of £320 – using MXN-GBP exchange rate for 2009 of 0.047476) and that this increase is almost exclusively due to Compartamos group lending. The households in treated neighbourhoods reported a growth in business size with both revenues and expenditure increasing. Whilst existing business grew in size, there was no effect on profits or on the probability of owning a business. There were no effects on labour supply or on income (apart from a decrease in income from government subsidies/aid). Overall household expenditure was unaffected, although the amounts spent on certain goods (e.g. temptation goods) went down. Finally, in terms of social and welfare outcomes, the surveyed women in treatment areas reported having more say in household decisions, having more trust in people and being happier.

What can we learn from this?

Overall, the study finds modest but mostly positive effects of the microcredit programme. The most significant effects are on business size and social/welfare outcomes. Given that there are some positive effects and little evidence of any negative effects the authors suggest that there is a policy case for expanding the use of such loans (e.g. by lowering interest rates). Furthermore, they suggest that further research on the heterogeneity of effects is necessary. If the modest average effects conceal the fact that some borrowers are made much better off and some worse off then there is an important effect on inequality that has been missed. It is important to note that these findings may not extrapolate from Mexico to the UK and it would be helpful to see this kind of trial replicated for similar UK lending schemes.


Angelucci, M., Karlan, D., and Zinman, J. (2014). Microcredit Impacts: Evidence from a Randomized Microcredit Program Placement Experiment by Compartamos Banco. [Study 542 in our Access to Finance Review, available here]

Read more case studies on innovation: Japan's emergency credit guaranteePartial credit guarantee programmeSee all


Related Content

Register for WWG Updates